This is a continuation of a 20-year program designed to understand leukemogenesis by the Abl oncogene and related issues. Because the gene is modular, we have focused on individual domains, most recently on SH3 because of its apparent anti-oncogenic properties. We plan to study the binding properties of SH3's from Abl and other proteins, using and comparing three quantitative methodologies. We will continue collaborative structural determinations and will use mutagenic methodologies to study the amino acids that determine specificity. We have found a number of binding sites for different SH3's, including Src, Crk, Nck, Grb2, Btk and Uau. In many cases, we have candidate molecules with SH3 binding sites and we want to put them into a biological content. We plan to characterize another modular protein element, the PH domain, which is found in many proteins. Here, as elsewhere, the yeast two-hybrid system will be an important tool. We also want to find new proteins that interact with Abl and to examine how Grb2 and Abl can synergize in cause cell transformation. We plan to study the Abl homologue in C. elegans to take advantage of the power of genetic analysis in that organism. Using a transient retrovirus packaging system we have recently developed, we plan to study further the tumors induced in mice by Bcr-Abl and other oncogenes, particularly the notch homologue TAN-1. With an amphotropic packaging line, we will examine the potential of using oncogenes to transform human cells. These studies can be expected to shed light on the causation of human chronic myelogenous leukemia by Bcr-Abl and on the general issues of the role of protein tyrosine kinases as signaling molecules in normal and tumor cells.